MC34063A 3V to 5V boost circuit

Hi there and a happy new year!

I want to design a circuit which boosts 3 to 5V, only at low current. For this purpose, I want to use a MC34063A.
The datasheet can be found here: http://www.onsemi.com/pub_link/Collateral/MC34063A-D.PDF
I also found a decent designing tool here: Nixie Tube Funkuhr - Nixieclock Schaltung und mehr...

But I have some questions, I hope you can help me:

1. I made an estimation for Vripple, about 100mV, is this realistic for two AA batteries as Vin?

2. I have no idea what I should enter for Fmin, some say 50kHz, some 100kHz (again, two AA batteries). What do you think?

3. Rsc is calculated to be 0.682ohms. How do I put this into practice?

4. Lmin is calculated to be 50uH. Does "min" mean I could also use a 100uH or so inductor?

Thank you!

I also found a decent designing tool here

Any web page that makes this statement should be regarded as suspect.

Typical input capacitor value is about 100uH,

Capacitors are not measured in uH, that is inductors.

I made an estimation for Vripple, about 100mV, is this realistic for two AA batteries as Vin?

The batteries have nothing to do with it, the ripple depends on current load.

I have no idea what I should enter for Fmin, some say 50kHz, some 100kHz

Again nothing to do with batteries. The higher the frequency the easier it is to smooth but the more demand is placed on the quality (ESR) of the capacitors.

Rsc is calculated to be 0.682ohms. How do I put this into practice?

Recalculate at a standard resistor value and see where that gets you.

Does "min" mean I could also use a 100uH or so inductor?

Yes.

Grumpy_Mike:

I also found a decent designing tool here

Any web page that makes this statement should be regarded as suspect.

I assume it just uses the formulas available in the datasheet, no magic there.

Grumpy_Mike:

Typical input capacitor value is about 100uH,

Capacitors are not measured in uH, that is inductors.

coule be just a typo?

Grumpy_Mike:

I made an estimation for Vripple, about 100mV, is this realistic for two AA batteries as Vin?

The batteries have nothing to do with it, the ripple depends on current load.

I know, but (I think) wont a battery have a higher Vripple than a lab power supply? But I realise this has also to do with the load, I'm going to experiment (and measure).

Grumpy_Mike:

I have no idea what I should enter for Fmin, some say 50kHz, some 100kHz

Again nothing to do with batteries. The higher the frequency the easier it is to smooth but the more demand is placed on the quality (ESR) of the capacitors.

Ok, here I understood something wrong. But how is this determinded? "Where" does it come from and what should I use for calculation?

Grumpy_Mike:

Rsc is calculated to be 0.682ohms. How do I put this into practice?

Recalculate at a standard resistor value and see where that gets you.

The nearest value would be 0.68ohms. But the cheapest resistor I found in this size-area is a 10W resistor with heat-sink which is $5 per part. Or do I search wrong?

Thanks

Ok, I noticed for Rsc I could just use parallel resistors, like 1ohm and 2.2ohm which wold give me 0.688 ohms in combination.

Oh and another one:

5. I need a 470pF and a 100uF capacitor, low-ESR. for the 100uF, I just look for a radial low-ESR electrolytic capacitor, but for the 470pF, will it a ceramic capacitor do?

Ceramic (cheaper and marginally better) and film capacitors will do.

If you really care for ripple, raise the output of your dc converter and follow-up with a linear regulator, or with a lc/rc network.

dhenry:
Ceramic (cheaper and marginally better) and film capacitors will do.

If you really care for ripple, raise the output of your dc converter and follow-up with a linear regulator, or with a lc/rc network.

Thank you, I just thought about a linear regulator, this will do the trick I think.

Ok, here I understood something wrong. But how is this determinded? "Where" does it come from and what should I use for calculation?

It is not determined. It is a design decision that you make. What frequency do you want to run the converter at?
Normally it is the highest possible but sometimes other considerations come into play.

wont a battery have a higher Vripple than a lab power supply?

No there is no ripple in a battery. However the ripple arises as a result of the up conversion.

Thank you very much, I think I understand now what you want to tell me.

Is the boost converter going to be permantly connected to the batteries?
The MC34063A has a minimum idle current of 4 ma so will flatten 2 AA batteries
in about 3 weeks , even with no load on the output.

mauried:
Is the boost converter going to be permantly connected to the batteries?
The MC34063A has a minimum idle current of 4 ma so will flatten 2 AA batteries
in about 3 weeks , even with no load on the output.

Yes, I know, thanks. The powerswitch will be in front of the chip.

in front of the chip.

Not sure what "chip" you are trying to power but chances are your avr will work happily at 3v or even lower, saving you all the trouble.

Another route would be to use a charge pump - they are far more efficient, particularly in space, if your load current isn't much.

A 555 timer can be turned into a quite efficient pwm controller, or even a single bjt + transformer.

dhenry:

in front of the chip.

Not sure what "chip"[...]

I was refering to the MC34063A.

Actually I need 5V to power a display.

Test drive it with these values

Ct=436 pF
Ipk=440 mA
Rsc=0.682 Ohm
Lmin=50 uH
Co=98 uF
R=180 Ohm
R1=1k R2=3k (5V)